Estimation Of Winter Wheat Residue Biomass Based On Full Polarimetric Radar Data

Wheat crop residues are left in the farmland of the stem portion of the crop after harvesting.Surface covered with crop residues holding soil moisture,carbon and other substances in arable land.Crop residues itself as a global carbon storage substance,it has a significant impact on carbon content globally calculation and evaluation of environment and climate.Rational use of crop residues can reduce the loss of carbon in agricultural soils,reduce greenhouse gas emissions and reduce surface soil organic matter losses caused by surface runoff.Therefore estimate agricultural crop residues biomass very important.In this paper,the study area is located in the North China Plain—Yucheng Experimental Zone.Using folly polarized Radar data(Radarsat-2)and Synchronization field observations data with radar data,carry on estimate winter wheat crop biomass residues study based on fully polarimetric radar data.Regression analysis of the observed data and winter wheat crop residue biomass,and study of characteristics between crop residues data and Radarsat-2 different polarization Backscattering coefficient.Then using a simplified Michigan Microwave Canopy Scattering(MIMICS)model simulations Backscattering coefficient of crop residues,and regression analysis of the simulations Backscattering coefficient and Radarsat-2 HH polarization mode Backscattering coefficient.Finally,build Winter wheat crop residue biomass estimation model based on Neural Networks,and using the measured samples crop residues biomass to evaluate the accuracy of simulation results.The main conclusion and results are as follows:1)Winter wheat crop residues biomass and winter wheat crop residues height has a good correlation.Correlation Analysis of wheat crop residue biomass data and crop residues height data,crop residue coverage data,soil moisture data at different depths and.The results show crop residues biomass data and crop residues height data has a good linear correlation(R2=0.73),and no significant correlation with crop residues coverage and soil moisture data at different depths.2)Winter wheat crop residues biomass data and HH polarization mode radar's backscatter coefficients has a good correlation.Analysis of the relationship between the Radarsat-2 HH,HV,VH,VV polarization mode scattering coefficient and the actual sample data.The results show it has a good correlation between residues biomass data and HH polarization mode radar's backscatter coefficients,the correlation coefficient R2=0.45,and no significant correlation with other polarization mode data.3)Simulation backscatter coefficients with the original backscatter coefficients has significant correlation.To determine canopy microwave scattering model—MIMICS model parameters,Simulation of wheat crop residues backscatter coefficients image.Comparison of the simulation backscatter coefficients image data with the original backscatter coefficients image data,the results show it has significant correlation(R2=0.78),so it can simulate wheat crop residues backscatter coefficient well.4)Derived wheat crop residues biomass has better accuracy based on MIMICS model.According to the actual measured Wheat crop residues vegetation parameters and Setting vegetation parameters and soil data fluctuation range,establish neural network estimation model estimate wheat crop residues biomass of the study area.According to field survey wheat crop residues biomass data,compared to derived wheat crop residues biomass data from the neural network estimation model,obtain accuracy assessment of derived wheat crop residues biomass,R2=0.588,the maximum relative error is 22.4%,and the average relative error is 7.13%.The results show it can estimate crop residues biomass according MIMICS model.